September 1, 2004
Volume 10, Issue 9 Abstract

CAG Repeat Length Differences Among a Racially Diverse Population of Prostate Cancer Cases

Thomas DE, Sankey SS, Neslund-Dudas C, Monaghan K, Tewari A, Rybicki BA


Dawn E. Thomas, MPH, Henry Ford Health System, 1 Ford Place, Detroit, MI 48202. E-mail:

Background: Excluding skin cancer, prostate cancer is the leading form of cancer among men in the United States. Several factors are thought to contribute to the risk of developing prostate cancer including genetic susceptibility, diet, and environmental exposure. In addition, the incidence of prostate cancer is higher in African Americans than Caucasians and African Americans often are diagnosed with a higher stage, grade, and prostate specific antigen (PSA). Reasons for this racial disparity are unknown, but genetics is thought to play a role. Studies have shown the androgen receptor CAG repeat length polymorphism to be associated with prostate cancer, with shorter repeat lengths increasing risk. African Americans tend to have shorter repeat lengths, which may help explain some of the racial disparity in prostate cancer.

Objective: To observe if there are differences in the androgen receptor CAG repeat lengths by race and other clinical variables.


Methods: Patients with newly diagnosed prostate cancer were enrolled in the Gene-Environment Interaction in Prostate Cancer study (GECAP) at Henry Ford Health System in Detroit, Mich. The study population had a diverse racial make-up of 50% African Americans (n = 77) and 50% Caucasians (n = 77). PSA level at diagnosis and Gleason score were collected on all cases. The androgen receptor CAG allele repeat length genotype was determined by polymerase chain reaction and analysis on an automated deoxyribonucleic acid sequencer. Statistical associations for continuous variables were assessed with Student's -tests. Chi-square tests were employed for categorical variables.






Results: At diagnosis, 25.7% of African Americans and 14.3% of Caucasians had a PSA level &#8805; 10 ng/dL ( = .10), 25.6% of African Americans and 20.1% of Caucasians had a Gleason score &#8805; 8 ( = .45). The average androgen receptor CAG repeat length for African Americans was 18.8 (SD, 3.6) versus 21.1 (SD, 3.3) for Caucasians ( < .01). Men with shorter CAG repeat length genotypes tended to have higher PSA levels at diagnosis ( = 0.30) and lower Gleason scores ( < .05).

Conclusion: Significant associations between the CAG repeat length genotype and both Gleason score and race was observed. Variation in the CAG repeat length genotype may help explain some of the racial differences in clinical presentation of prostate cancer we observed in our study population, as well as the overall racial disparity in prostate cancer incidence.

Validity of a Family History Form as a Genetic Screening Tool

Dobin S, Saucier J, Whetteckey J, Rajab MH


Mohammad H. Rajab, PhD, MPH, Scott & White Memorial Hospital, Texas A&#38;M University, 2401 S 31st Street, Temple, TX 76508. E-mail:

Background: In this era of advanced molecular genetics, genetic screening is becoming a vital component of medicine. Primary physicians need screening tools that will allow them to take advantage of this genetic revolution.

Objective: To validate a locally developed family history form (FHF) as a genetic screening tool for primary physicians. If successful, the FHF was to be implemented by our health maintenance organization for use by primary physicians.

Methods: The FHF was designed based on contributions from physicians, biostatisticians, and FHFs used by other institutions. The design was a bubble form with a list of approximately 50 genetic disorders. Twenty patients, seen in the genetic clinic, consented to participate in this study. Participants were asked to fill out the FHF at the beginning of their appointment. A genetic counselor then obtained a 3-generation pedigree from each patient. The FHF was considered valid if it agreed with at least 90% of the family history information gathered by the pedigree.


Results: Of the 20 FHFs completed, 4 matched at least 90% of the pedigree information ( > .9). The overall percentage of agreement for all diseases and for all patients was 56.4%. The information that was captured by the pedigree and not by the FHF included half-siblings, third-degree relatives, carrier status, and ages at diagnosis. In 14 FHFs, the form picked up some family information that was not obtained by the pedigree. This unexpected finding is most likely due to the indication for their genetics appointment influencing the pedigree interview.

Conclusion: We do not recommend the use of a bubble FHF design as a genetic screening tool, and we plan to test different designs including fill-in-the-blank.